RISKeLearning Nanotechnology Applications and Implications for Superfund

Transcription

1 RISKeLearning Nanotechnology Applications and Implications for Superfund March 15, 2007 Session 3: Nanotechnology DNAPL Remediation Dr. Matt Hull, Luna Innovations, Inc. Dr. Peter Vikesland, Virginia Tech Dr. Greg Lowry, Carnegie Mellon University Organizing Committee: SBRP/NIEHS EPA MDB William Suk Michael Gill Nora Savage Maureen Avakian Heather Henry Jayne Michaud Barbara Walton Larry Whitson Claudia Thompson Warren Layne Randall Wentsel Larry Reed Beth Anderson Marian Olsen Mitch Lasat Kathy Ahlmark Charles Maurice Martha Otto 1

2 Magnetite Nanoparticles for Remediation of Contaminated Groundwater Funded by an EPA SBIR grant (Contract No. EP-D ) Matthew Hull Principal Investigator Luna Innovations Incorporated Blacksburg, VA Peter Vikesland, Ph.D. Assistant Professor, CEE Virginia Tech Blacksburg, VA Copyright Luna Innovations

3 Overview EPA SBIR Program Magnetite Characterization and Reactivity Why Magnetite? Synthesis and Characterization Carbon Tetrachloride Preliminary Results Encapsulation of Nanoparticles Thoughts on Nanotechnology EHS Conclusions 3 Copyright Luna Innovations

4 EPA SBIR Program Goal: Research, develop, and commercialize high-risk/high-payoff technologies that help solve environmental challenges. Phase I Demonstrate Technical Feasibility Phase II Optimization and Scale Up Phase III Commercialization of Actual Product 4 Copyright Luna Innovations

5 Luna/VT EPA SBIR Program (Status: Phase II) Develop magnetite nanoparticles for remediation of chlorinated organic compounds (other targets) Develop delivery strategies that facilitate particle delivery to subsurface and interaction with target Scale-up production of Phase II-optimized magnetite form for commercial application 5 Copyright Luna Innovations

6 Magnetite Characterization and Reactivity 6 Copyright Luna Innovations

7 ZVI Permeable Reactive Barriers Effectively Treat Many Groundwater Contaminants Fe(0) Fe(II) Fe(III) Nanoscale corrosion products form over time Magnetite Nanocrystals Secondary precipitates Iron metal (Source: Kohn, Livi, Roberts, and Vikesland, ES&T 2005) 7 7

8 Why magnetite? Biogenic Magnetite Magnetotactic Bacteria McCormick and Adriaens, 2002 Magnetite shell On RNIP (Fe H2 ) MRI Contrast Magnetite Nanocrystals Nurmi et al., 2005 Kohn et al,

9 Magnetite Synthesis and Particle Characterization Co-precipitation Method Vayssières et al. J. Coll. Int. Sci Mixture of FeCl 3 and FeCl 2 to NaOH Rapid stirring IEP: 6.5 ± 0.25 Mean Particle Dia. = 9.2 ± 1.6 nm 100 nm Synthesis done under conditions of strict oxygen exclusion 9 9

10 Carbon Tetrachloride (CT) Manufactured Chemical Refrigeration fluid Propellants for aerosol cans, fire extinguishers Pesticide Cleaning fluid, Degreasing agent CT and Human Health Found at 425 of 1,662 EPA NPL sites Exposure via contaminated air, water, soil Damage to liver, kidneys, nervous system. Likely Carcinogenic (DHHS, IARC) EPA limit for drinking water 5 ppb Difficult to Remediate Generates toxic intermediates: Chloroform (CHCl 3 ) Dichloromethane (CH 2 Cl 2 ) Current methods transfer pollution problem Source: CT Breakdown Pathways :CCl: 2 CCl 4 CCl 3 HCOOH Copyright Luna Innovations 2006 CO CHCl 3 CH 2 Cl

11 Carbon Tetrachloride Degradation by Magnetite Reaction rates vary with the size of the particles CCl 4 (μm) k obs,ct = h -1 k m = Lg -1 h -1 k SA,BET = ( ) x 10-5 Lm -2 h -1 k obs,ct = h -1 k m = Lg -1 h -1 k SA,BET = ( ) x 10-3 Lm -2 h -1 <80 nm> magnetite <9 nm> magnetite k obs = k SA [Fe 3 O 4 (m 2 /L)] k m ρ m = k SA a S ρ m ph 7.8, 5 g/l Fe 3 O 4, 50 mm HEPES, 0.1 M NaCl Time (hrs) 11 Reporting km values not ksa; ph range was limited by HEPES buffer 11

12 Commercially Available Magnetite Not Reactive Three commercially available magnetite products Concentration (μm) A 11 g/l g/l 1530 B 1 g/l 1 g/l 1530 B 3 g/l 3 g/l 9002 C 1 g/l 1 g/l 9002 C 3 g/l 3 g/l Variable loading 1 g/l vs. 3 g/l Essentially no reactivity observed Time (hrs) 12 Copyright Luna Innovations

13 Rxn Rate for CCl 4 Reduction Increases from ph 6.1 to ~ 8 k m (Lg -1 h -1 ) <9 nm> magnetite; M NaCl <9 nm> magnetite; 0.1 M NaCl <9 nm> magnetite; mass loading experiment; M NaCl <9 nm> magnetite; mass loading experiment; 0.1 M NaCl <80 nm> magnetite; 0.1 M NaCl Danielsen and Hayes (2004); 0.1 M NaCl k m for NZVI ~ Lg 1 h -1 (Nurmi et al. 2005) k m for bulk magnetite is consistent with previous study ph 13 13

14 Surface Area Normalized Rate Constants g/l Experiment BET Normalized 5 g/l Experiment TEM Normalized Danielsen and Hayes (2004) Log(k SA (Lh -1 m -2 )) } Size effects? ph

15 Ionic Strength Affects Reduction of CCl 4 CT degrades faster at lower ionic strength M k m (Lh -1 g -1 ) NaCl Z avg (nm) M 0.01 M M Ionic Strength Time (min) ph 7, 50 mm HEPES, [CT] 0 = 100 μm, 5 g/l magnetite Stable aggregate size measured using Dynamic Light Scattering 15 15

16 Ionic Strength Affects Reduction of CCl 4 CT degrades faster at lower ionic strength Effect is magnified at higher ph values k m (Lh -1 g -1 ) NaCl Ionic Strength ph 7, 50 mm HEPES, [CT] 0 = 100 μm, 5 g/l magnetite ph 7.8, 50 mm HEPES, [CT] 0 = 100 μm, 5 g/l magnetite 16 16

17 Ionic Strength Effect on Colloidal Stability M NaCl 0.01 M NaCl Low ionic strengths After 1 hour As particles approach each other the electric double layers overlap and cause repulsion High ionic strengths In increased ionic strength solutions the electric double layers shrink thus allowing the particles to get closer without being repelled 17 Magnetic forces. 17

18 Degradation Rate Affected by Electrolyte Identity k obs (hr -1 ) NaClO 4 NaCl CaCl 2 The rate of loss of CCl 4 in presence of CaCl 2 was much lower than in presence of NaCl or NaClO 4 The effect of electrolyte identity becomes less important at higher salt concentration e Ionic Strength (M) The rate of loss of CCl 4 in M NaClO 4 was faster compared to NaCl 18 18

19 Magnetite Reactivity 19 Reactivity may be attributed to octahedral layer in its inverse spinel crystal structure 19

20 Chloroform Yields Increase with ph ph effects mass loading NaCl Ionic strength CaCl2 Ionic strength CF yield increases with a decrease in ionic strength CF final /CT initial ph 20 20

21 Magnetite is Topotactically Converted to Maghemite Fe 3 O 4 γ Fe 2 O Anoxic magnetite (sonicated) Aerated magnetite (sonicated) Z avg (nm) Time (min) Aggregate restructuring may occur during oxidation 21 21

22 Aggregation Behavior may be Affected by Oxidation Fe 2+ Fe 3 O 4 Fe 3 O 4 γ Fe 2 O 3 γ Fe 2 O

23 Thoughts on CCl 4 Reactivity Fe 3 O 4 + CCl 4 Æ γ-fe 2 O 3 + Fe 2+ + CHCl 3 Fe 2+ Z avg Fe 3 O 4 Z avg,t=0 Possible rate modifying effects: 1. Diffusion of Fe 2+ within a magnetite particle 2. Diffusion of CCl 4 to a reactive site within aggregate or on surface of aggregate 3. Electron transfer 23 23

24 Encapsulation of Nanoparticles for Environmental and Biological Applications 24 Copyright Luna Innovations

25 Concept and Advantages Maintain anaerobic microenvironment Trojan Horse Concept Microvessels for addition of rateenhancing reactants, remediation cocktails (Quinn et al., 2005) Functionalize surface for enhanced suspension properties May facilitate long-term storage under ambient conditions Add control over reaction chemistry, kinetics Source: Encapsulation affords an added layer of engineering control over reaction chemistry and timing 25 Copyright Luna Innovations

26 Precedent: NASA EZVI (Quinn et al., 2005) Two components: ZVI for reductive dechlorination Veg oil for enhacement of microbes EZVI droplet: Oil-liquid membrane surrounding particles of ZVI in water. Unclear whether reactivity with TCE due to ZVI or microbial enhancement Injection methods can damage EZVI droplets (Source: Quinn et al, ES&T 2005) 26 26

27 Encapsulation Protective Layer Iron Demonstrated with nano- to micro-scale particles, in-house and commercial formulations Magnetite-filled Capsule 27 Copyright Luna Innovations

28 Encapsulated Magnetite Nanoparticles Beta Testers, Product Development Partners Welcomed 28 Copyright Luna Innovations

29 Burning Questions Reactivity of encapsulated magnetite nanoparticles? Control/tunability of the capsule/particle composite? Preservation of particle reactivity in the capsule? Breakdown and particle release process? How much DNAPL does it take to get to the center of the ironfilled capsules? 29 Copyright Luna Innovations

30 Thoughts about Nanotech EHS Realize societal benefits of nanotechnology responsibly Unique reactivity Possible unintended effects Luna NanoSafe : Started in 2003 to address EHS concerns proactively Q2 2007: Begin third party ecotox testing of various nano iron species and composites Adapted From Mark Alper, DOE Molecular Foundary LBNL 30 Copyright Luna Innovations

31 Conclusions Magnetite prepared under strict anaerobic conditions can degrade CT without adding Fe II. On a mass basis, CT loss rates are enhanced in the presence of nanoparticulate magnetite. Chloroform yields are affected by both solution ph and ionic strength. Surface normalization of the reactivity of nanoparticle suspensions may be unwise. Encapsulation may offer additional engineering control over interaction of particles with remediation target. 31 Must determine EHS issues with nano iron Copyright Luna Innovations

32 Acknowledgments Funding US EPA SBIR Program NSF (Dr. Vikesland s Lab) Virginia Tech Environmental BioNanoTechnology Lab April Heathcock Erik Makus Rob Rebodos John Templeton Luna Innovations Life Sciences Group Len Comaratta Steven Abbott Natasha Belcher Advanced Materials Group Kristen Selde Bryan Koene 32 32

33 Contact Information Matthew Hull Principal Investigator Phone:

34 Functionalized Reactive Nanoscale Fe 0 (NZVI) for in situ DNAPL Remediation: Opportunities and Challenges Gregory V. Lowry Associate Professor of Environmental Engineering Carnegie Mellon University, Pittsburgh, PA , USA NIEHS Webinar Series on Nanotechnology March 15, R

35 Students and Collaborators Faculty Robert Tilton (Chem Eng) Krzystzof Matyjaszewski (Chemistry) Post Docs and Students Dr. Abdulwahab Almusallam, Dr. Bruno Dufour, Dr. Jeongbin Ok, Dr. Traian Sarbuu Dr. Yueqiang Liu, Tanapon Phenrat, Navid Saleh, Kevin Sirk, Hye-Jin Kim Dan Schoenfelder 35 35

36 DNAPL Contamination DNAPL SOURCE ZONE TREATMENT Plume Concentration Δt MCL Nanoiron Injection Δm How many injections? How long? NO SOURCE POST ACTION REMOVAL REMOVAL Mass emission is a function of the total DNAPL mass and architecture. Reducing the source mass (Δm) decreases mass emission and downgradient loading. Cost-effectiveness relies on effective placement of nanoiron

37 Treatment with Reactive Nanoparticles Effective Treatment Requires 1. Mobility 2. Target specificity 3. High reactivity and long lifetime 4. Minimal risk of nanoparticle migration to sensitive receptors (low concentration) 37 37

38 Optimizing Two Scenarios High concentration of Nanoparticles At injection site Need to maximize mobility to be able to deliver the materials Optimize remediation performance Low concentration of Nanoparticles After dilution in aquifer Need to minimize mobility to ensure that nanoparticles remain in place Minimize risks 38 38

39 Functionalized Reactive Nanoiron (NZVI) Nanoiron (RNIP) Surface modifiers + PSS PAP TCE + Fe 0 Æ HC Products + Cl - + Fe 2+ /Fe 3+ Liu, Y., Lowry, G.V. et al, (2005) ES&T 39, 1338 Liu and Lowry (2006) ES&T 40, 6085 p m n O O O O H SO 3H M n=2000 M n=5700 M n=8340 PMAA-PMMA-PSS RNIP Modifier ζ p potential (mv) Average Dia (nm) RNIP (none) -29.6± ±4 PMAA 48 -PMMA 17 -PSS ± ±21 SDBS ± ±15 15 MRNIP (PAP MW=2.5k) -37.6±1.1 66±3 PAP (MW=2.5k) -51.7± ±18.6 PSS (MW=70k) -48.9± ±

40 Outline Getting particles to DNAPL (NP mobility) Aggregation Attachment/filtration Coatings to minimize filtration DNAPL degradation rates (Reactivity and Lifetime) Effect of groundwater geochemistry and surface coatings Fate of particles and coatings DNAPL Targeting Approaches for in situ targeting 40 40

41 Nanomaterial Mobility in Porous Media A---Aggregation Aggregation B---Straining C---Attachment D---NAPL Targeting Factors affecting mobility Chemical ph, I, surface chemistry Physical flow velocity, particle/aggregate size, heterogeneity Lowry, in Env. Nanotech. (in press)

42 Nanoparticle Aggregation Nanoparticles aggregate in water: High Hamaker constant-i.e. attractive van der Waals forces Chemical bonding Hydrophobicity Magnetic attraction (Fe 0 ) Nanoparticles have high diffusion coefficients and many particle collisions 42 42

43 How Long is Bare NZVI Nano? Aggregate size depends on Particle concentration Time Magnetic properties M s RNIP >M s Magentite >M s Hematite Phenrat T., Lowry G.V., et al., (2007) ES&T 41,

44 NZVI Aggregation & Sedimentation Φ=10-5 (~80 mg/l) 1-min Nanoiron sedimentation curves (1 mm NaCl) 25 micron 9-min 25 micron 35-min 25 micron ~ micron diameter (D F =1.8) Phenrat T., Lowry G.V., et al., (2007) ES&T 41,

45 Attachment to Surfaces Attachment is an important fate process Limits mobility in porous media May limit bioavailability/transformation/degradation Attachment is a function of the particle and coating type Differences between een NPs and coatings 45 45

46 Attachment Limits Mobility Inlet Time=1 min Monolayer of sand 1 1/2 Micro-fluidic PDMS cell Outlet Nanoiron aggregates are filtered 26 μm Time=10 min 26 μm Saleh, N. Lowry, G.V. et al. Environ. Eng. Sci. 24 (1) 2007 p

47 Surface Modifiers Inhibit Aggregation and Attachment and Increase Mobility Common Surface Coatings Polymers/Polyelectrolyte 9Triblock copolymers (electrosteric) 9Polyaspartic acid (electrostatic) 9Cellulose/polysaccharides (steric) 9PEG (steric) Surfactants 9SDBS (electrostatic) Inhibits Aggregation - Inhibits Particle Media Interactions - - Charge Stabilization water Steric Stabilization mineral surface 47 47

48 Copolymer (MW=125k) PMAA 48 -PMMA 17 -PSS 650 Different Modifiers Increasing MW Polystyrene sulfonate (MW=70k) Polyelectrolytes Polyaspartic acid (MW=3k and 10k) Surfactant C 12 H 25 (C 6 H 4 )SO 3-48 SDBS (MW=350) 48

49 Modifiers Inhibit Agg/Sed Largest Polymer Least aggregation No Polymer Most aggregation Saleh, N., Lowry, G. V., et al. (2005). Nano Lett. 5 (12) Saleh, N., Lowry, G. V., et al. Environ. Eng. Sci. 24 (1) 2007 p

50 Modifiers Decrease Attachment to SiO 2 Surfaces Sand Grain Sand Grain QCM is a good predictor of Saleh et al. Environ. Eng. Sci. 24 (1) 2007 p mobility! 50 50

51 Modifiers Enhance RNIP Mobility 9 All modifiers enhance mobility relative to bare RNIP 9 Variation between L=12.5-cm silica sand column with porosity of Particle polymers and surfactants implies potential to select a Kogyo, Inc. The approach velocity was 93 m/d. transport distance concentration is 3 g/l and I=1mM. Modifying agents were added at 2g/L concentration in each case. MRNIP was supplied by Toda Saleh et al., 2007 EES 24(1)

52 Mobility Depends on Ionic Strength and Composition Na + inhibits mobility Ca 2+ inhibits mobility more than Na + Saleh, N. et al. ES&T (in prep) Sand L=61 cm porosity=0.33 Velocity 3.2x10-2 cm/s I= mm Na + or Ca mg/l particles 52 52

53 Applying a simple filtration model yields the predicted transport distance needed for 99% removal Mobility depends on Site Geochemistry Modi difier Na + (mm) Log α (--) Dist. (m) Ca 2+ Ca 2+ (mm) Log α (--) Dist. (m) Poly lymer (MW=1 =125k) Aspartate (MW=3k =3k) SDBS (MW=3 =350) Site K + + Na + mm Ca 2+ + Mg 2+ mm Alameda a Point, CA Paris Island, SC Mancelona, MI Typical concentrations of monovalent and divalent cations 53 53

54 Real Sites are NOT Sand columns! Scales of Interest Field-Scale IncreasingComplexity Intermediate-Scale Column 2-d cell Need to Up-Scale process parameters determined in the laboratory to field scales 54 54

55 NZVI Reactivity and Lifetime Fundamental Questions What are the reaction r r rates and products? How long do the particles remain r reactive? What geochemical factors affect their reactivity and lifetime? How do surface e modifiers affect reactivity? Liu et al, (2005) ES&T 39, 1338 Liu and Lowry, (2006) ES&T, 40 (19) 6085 Liu, Phenrat, and Lowry, (2007) ES&T, (in prep) 55 55

56 Reactive Fe 0 Nanoparticles Contaminants are reduced Nano Fe 0 is oxidized Fe 0 TCE Fe 3 O 4 Fe 0 Fe 0 Fe 3 O 4 Acetylene + Fe 3 O 4 Ethene + Lifetime depends on Oxidant loading and ph Ethane H 2 H + H + is reduced 56 Liu et al, (2005) ES&T 39, 1338 Liu and Lowry, (2006) ES&T, 40 (19)

57 Types of Nanoiron Borohydride reduction 1,2 2Fe 2+ + BH H 2 O Æ 2Fe 0 + B(OH) 3 + 2H 2 + 3H + Gas phase reduction by H 1 2 Toda Kogyo (RNIP) FeSO 4 Æ FeO(OH) H2 Æ Fe 0 H 2 O Fe 3 O 4 shell Fe 0 core ~5000 ppm S (reduced) 1 Liu, et al., Environ. Sci. & Technol. 2005, 39, Liu et al., Chem Mater. 2005,

58 Differences in Reactivity Fe(B) RNIP Faster Reaction k=1.4 X 10-2 L hr -1 m -2 Saturated Products TCE t 1/2 =~ 2 hr (@2g/L) *Iron Filings k=10-3 to 10-4 L hr -1 m -2 Slower reaction k=3 X 10-3 L hr -1 m -2 Unsaturated Products TCE t 1/2 =~ 8 hr (@2g/L) 1 Liu, et al., Environ. Sci. & Technol. 2005, 39,

59 NZVI Lifetime Depends on ph Fe 0 + 2H + Fe 2+ + H 2 ~2 weeks ph=6.5 ~9 months ph 8.0 Liu and Lowry, (2006) Environ. Sci. Technol., 40 (19)

60 TCE Dechlorination Over Particle Lifetime 11 k obs,tce is relatively constant as particles age Liu and Lowry, (2006) Environ. Sci. Technol., 40 (19)

61 Effect of TCE Concentration on Fe 0 Utilization Small effect of TCE Use in source zone concentration on increases Fe 0 reactivity efficiency 61 Liu, Phenrat, and Lowry, Environ. Sci. Technol., (in prep) 61

62 Effect of Groundwater Solutes on Reactivity with TCE Dissolved solutes had no effect on H 2 evolution Dissolved solutes lower reactivity by a factor of 2 to 7 depending on the solute Effect follows expected trend of solute strength of complexation with HFO Liu, Phenrat, and Lowry, Environ. Sci. Technol., (in prep) 62 62

63 Effect of Modifiers on RNIP Reactivity with TCE PMAA 48 -PMMA 17 -PSS 650 modified RNIP: 10 times less reactive then unmodified RNIP, but still reactive enough TCE t 1/2 6 days (at 2 g/l) for the lowest activity modified particles Saleh, G. V. Lowry, et al. Environ. Eng. Sci. 24 (1) 2007 p

64 Contaminant Source Zone Targeting Contaminant Source Zone Without Targeting Nanoparticles can flow past source zone Nanoparticle surface coatings can provide an affinity for DNAPL 64 64

65 Potential Strategies for Targeting Interfacial targeting Destabilization Targeting Controlled placement O H O Hydrophobic n O O m SO 3 H M n =2000 M n =5700 M n =8340 Hydrophilic p Particles Attach to the NAPL/water interface Saleh, N., Lowry, G. V., et al. (2005). Nano Lett. 5 (12) water solvent 65 65

66 Interfacial Targeting Challenges TCE Nanoiron Trajectory at different porewater velocities Water Flow velocities: μm/s (2.6-13m/day) Residence time: 1-10 s Baumann, T., Keller, A. A., Auset-Vallejo, M., Lowry, G V. (2005). AGU Fall Meeting, San Francisco, CA, December 5-9,

67 Destabilization Targeting Polymer adsorption to RNIP is strong and effectively irreversible Higher MW=stronger sorption Mitigates concern of NAPL mobilization Modi difier Initia ial Adso sorbed mass (mg/m 2 ) Percent Remaining g Adsorbed 2weeks 4weeks 8weeks PAP 2.5K 0.85± ± 3 86 ± ± 5.5 PAP 10K 1.47± ± ± ± 2 2weeks 5weeks 8weeks PSS 70K 2.89± ± ± ± 0.6 PSS 1M 2.55± ± ± ± 4.7 2weeks 6weeks 8weeks CMC 90K 2.09± ± ± ± 3.1 CMC 700K 3.71± ± ± ± 0.9 Kim, H-J., Lowry, G.V. et al., (in prep) 67 67

68 Strategies for Controlled Placement of Nanoiron Geochemical conditions change from the injection well down gradient due to dilution. Potential geochemical changes that can afford targeting include: Ionic strength variation (from low to high) Velocity variation (from high to low) Partial saturation DNAPL saturation varies from saturated at a pool surface to just a few percent at the fringe. DNAPL architecture may afford targeting opportunities Hydrodynamic trapping Co-solvency effects Saturated DNAPL 68 68

69 Conclusions Aggregation and attachment limits bare Fe 0 NPs mobility in aquifers Surface modification increases mobility GW geochemistry (Ca 2+ and Mg 2+ ) controls mobility Mobility of 10 s of meters possible with appropriate coatings at typical GW ionic composition NZVI is highly reactive with TCE under environmental conditions Lifetime depends of geochemistry and oxidant loading Use in NAPL source zone maximizes Fe 0 utilization In situ targeting of entrapped NAPL r requires optimization of the coatings Matching modifier and GW geochemistry offers potential for controlled placement 69 69

70 Nano vs. Micro Greater surface area of nano ( m 2 /g) provides higher reactivity than micro (~0.1 m 2 /g) nanoiron Æ 0.5 to 1.5 lb/yd 3 ; microiron Æ > 20 lb/ydd 3 Delivery to source Nanoiron-direct - push wells Micoiron-high - pressure injection and greater cost Total cost includes Management/engineering Injection services Materials (~15% at pilot scale) 70 70

71 Field Validation Needed Assessment of the effect of treatment ent on the DNAPL mass and mass emission from the source is needed Pilot-scale field demonstration WITH substantial characterization acterization before, during, and after treatment is needed Better understanding of the between NZVI and the microbial communities at a site are needed 71 71

72 Acknowledgement Toda Koygo Corp. U.S. EPA-STAR (R830898) US DOE EMSP Program (DE-FG07 FG07-02ER63507) US DoD (SERDP) (W912HQ-06-C-0038) 0038) US NSF (CBET ) R

73 Thank You After viewing the links to additional resources, please complete our online feedback form. Thank You Links to Additional Resources